What is Qwen Code?
Qwen Code is an open-source AI coding agent for the terminal from the Qwen team at Alibaba. It installs a command called qwen, reads your project, edits files and runs commands from plain-language instructions. You choose how it authenticates — and which model provider it uses — from inside the tool.
Tallos is a desktop agent development environment that installs Qwen Code for you, launches it in isolated workspaces and runs it next to Claude Code, Codex, Gemini CLI and your other agents.
How do I set up Qwen Code in Tallos?
Install Qwen Code from Settings → Connections, sign in from a Qwen Code session, and set it as your default agent. If qwen is already on your PATH, Tallos detects it.
- 1
Install
Click Install on the Qwen Code card. Tallos shows the official command before running it:
brew install qwen-codeornpm install -g @qwen-code/qwen-code@lateston macOS, andnpm install -g @qwen-code/qwen-code@lateston Windows (Node.js required). - 2
Sign in with /auth
Click Sign in. Tallos opens
qwenin a terminal; type/authin the session and pick your provider. Qwen Code saves that choice in its own settings. - 3
Make it the default
In Settings → Agents, set Qwen Code as the Default agent so it is pre-selected in every new workspace.
- 4
Start a task
Add a repository, click + for a workspace and describe the task. Tallos creates a git worktree on its own branch, starts
qwenthere and types your prompt in once the session is up.
Walk through every agent on the page in connect your agents, or read how agent connections work.
How do I run multiple Qwen Code sessions in parallel?
Open one workspace per task. Each workspace is a separate git worktree with its own branch, files and terminals, so several Qwen Code sessions can work on the same repository without touching each other's files.
- One branch per task. Installs, builds and tests in one session never leak into another.
- Side by side. Split panes in any direction to watch several Qwen Code sessions at once.
- Review per branch. Each workspace has its own diff; comment on lines and send the feedback back to that session.
- Merge only what you approve. When a branch is good, commit, push and open a pull request from Tallos. Delete the workspaces you don't need.
See parallel workspaces for how the isolation works.
Install Qwen Code and run it in parallel workspaces.
How do I use Qwen Code alongside other agents?
Put it in a squad — a leader agent plus members, each in its own worktree — or run it next to other agents in separate workspaces. Any detected agent can fill a slot, Qwen Code included.
- As a member. Describe its job in plain words, set How many copies you want, and the leader assigns work to each one.
- As one of three attempts. In *3 attempts, pick the best*, Qwen Code can try the same objective as two other agents; the leader recommends one result.
- As the leader. It plans, dispatches and reports within your limits — 5 rounds and 4 members at once by default, up to 20 rounds and 12 members.
Tallos does not pass a model flag to Qwen Code, so a Qwen member runs the model set in Qwen Code's own configuration. Nothing merges without your approval the squad's result.
Does Qwen Code open in a chat view?
No. Qwen Code opens in the terminal view, with its own interface. The chat view is available for agents whose transcripts Tallos can read, such as Claude Code, Codex and Grok CLI.
What does Tallos add to Qwen Code?
| Capability | Qwen Code on its own | Qwen Code in Tallos |
|---|---|---|
| Parallel sessions | Open more terminals and set up git worktrees yourself | One workspace per task, each with its own worktree and branch |
| Install | Run the install command yourself | One click on the Connections card, with the command shown first |
| Sign-in | /auth inside a session | The same /auth, opened from the card's Sign in button |
| Model choice | Set inside Qwen Code | Unchanged: Qwen Code uses its own configured model |
| Review | Your own git tools | Diff per workspace, line comments sent back to the agent |
| Other agents | Separate tools | Same app; squads with a leader, members and best-of-N |
Frequently asked questions
Do I need a separate subscription to use Qwen Code in Tallos?
No. Qwen Code uses the provider and account you set up with /auth, and Tallos does not resell model access.
Why can't I paste an API key for Qwen Code in Tallos?
Qwen Code only uses an API key after its own settings declare the provider it belongs to. Configure the provider inside Qwen Code with /auth instead.
Why does Tallos show Qwen Code as 'sign-in unknown'?
Qwen Code has no command that reports sign-in status from outside a session. Run /doctor inside Qwen Code to check your setup.
Can I pick Qwen Code's model from Tallos?
No. Tallos can't list Qwen Code's models, so choose the model inside Qwen Code; every session and squad member uses that configuration.
Does Qwen Code work in Tallos on Windows?
Yes. On Windows 10+, the Connections page installs Qwen Code with npm, so install Node.js first.
Does Tallos change Qwen Code's approval mode?
Only if you ask it to. By default Tallos starts Qwen Code in manual mode, so its approvals stay on. If you choose the autonomous mode in Settings → Agents → Agent Permissions, Tallos launches it with --approval-mode yolo.
